Research On Warehouse Location Method Based On Passive RFID

With the rapid growth of China’s economy and the rapid development ofe-commerce, the scale of the logistics industry is constantly expanding. Warehousingas an important part of the logistics process, its management efficiency directlyaffects the efficiency of the entire logistics proces, thus affecting the efficiency of theentire supply chain. In the warehouse environment, cargo’s real-time location is thepremise of various operations in the warehouse management, which can effectivelyimprove the efficiency of warehouse mamgement.Radio Frequency Identification (RFID) is an automatic identification and datacapture technology. With the characteristics of multi-target reading and writing,remote identification and reusable, it is ideal for wireless location-based applications.In this thesis, we mainly focus on the research of passive RFID’s location methodsand applications in the warehouse. Through the in-depth study of RFID system’sworking principle and characteristics, location algorithms based on RFID aresummarised and classified. Combined with the actual analysis of a large warehouse,the requirements of warehouse positioning system are summarised, and on this basis,the architecture of intelligent warehouse positioning system is proposed which ismainly focusing on the positioning scheme of store shelf. Based on the proposedpositioning scheme of store shelf, the two-step adaptive reference tag positioningalgorithm is proposed. Finally, through the simulations and experiments based onhardware and software platforms, performace of different positioning algorithm arecompared and future research directions are discussed. The main work of this paper isas follows:(1) On the basis of in-depth study of RFID technology, this paper makes acomprehensive study of indoor location technology, especially emphasizing on theRFID-based indoor positioning technology, which can be divided into triangulation,scene analysis and proximity. Menwhile, two classical RFID location systemsLANDMARC and VIRE are introduced, which are the research basis of this paper.(2) Design the architecture of warehousing positioning system, including the overall architecture of the positioning system, the architecture of the physical layerand different positioning strategies, specific positioning schemes of different storageareas are also introduces briefly.(3) With the combination of LANDMARC and VIRE, this paper proposes atwo-step adaptive reference tags positioning algorithm. Through the combination ofsignal strength and read counts, the target location region is divided into small parts,coarse-grained positioning is applied first and targets are located to one of the dividedsmall regions, then fine-grained positioning is applied and tergets are located moreprecisely, thus the efficiency is improved greatly. Meanwhile, in order to solve theproblem of lacking reference tags on the boundary, boundary tag expansion areadopted to improve the location accuracy of the tags locating on the boundary. Finally,positioning error elimination methos is introduced, and positioning errors areeliminated to some extent.(4) Through the building of software and hardware experimental platform, theperformance of the proposed algorithm is evaluated. According the data collectedfrom the actual experiment platform and the model established, performance ofseveral localizaitoning algorithms is analysed and compared based on simulation.Meanwhile, four important factors which affect the localizaiton performace arefocused specially, including the reference tag density, the number of antennas, thenumber of nearest reference tags and density of cirtual reference tags.Experimental results show that the proposed algorithm has a smaller averagelocation error and maximun positioning error, and positioning solution shows goodstability, the overall performance of the proposed algorithm is the best.